    s : dict
        Settings dictionary
    df_set : pd.DataFrame
        Dataframe containing the list of proteins to process, including their TMD sequences and full-length sequences
        index : range(0, ..)
        columns : ['acc', 'seqlen', 'TMD_start', 'TMD_end', 'tm_surr_left', 'tm_surr_right', 'database',  ....]
    logging : logging.Logger
        Python object with settings for logging to console and file.
    acc : str
        Protein accession number. (e.g. UniProt acc, PDB_acc + chain letter)
	acc : str
        Protein accession (e.g. UniProt, PDB)
    database : str
        Database name, e.g. "crystal", "NMR" or "ETRA".
    TMD_seq : str
        TMD sequence
    TMD_seq_pl_surr : str
        TMD sequence plus surrouding residues (usually 20) for BLAST
    blast_xml_file : str
        Path to xml file with BLAST results from NCBI.
    xml_txt : str
        Path to a text file that saves the download date, etc.
    xml_tar_gz : str
        Path to compressed tar file containing blast_xml_file, with BLAST results from NCBI.
    expect_value : float or int
        BLAST parameter "expect", which gives the expected number of random matches after BLAST with these parameters.
        Typically 10.
    hit_list_size : int
        Maximum number of BLAST hits.
    train_set_list : list
        List of training datasets in selection
        E.g. ["set02", "set04"]
    test_set_list : list
        List of test datasets in selection
        E.g. ["set03", "set31"]
    mult_THOIPA_dir : str
        Path to folder containing results for multiple THOIPA comparisons.
    mult_testname : str
        String denoting this combination of test and training datasets
        E.g. testsets(2)_trainsets(2)
    sheet_name : str
        Excel sheet_name
        This is the toggle deciding whether the OLD or NEW performance measure is used
        Default = new method ("df_o_minus_r"), where the overlap MINUS random_overlap is used
    suffix : str
        Suffix for figure
        E.g. "" or "_old_method_o_over_r"
	
    THOIPA_pred_csv : csv
        THOIPA result for this testset-trainset combination
        Columns = "residue_num", "residue_name", "THOIPA"
        Index = range index of residues
    combined_incl_THOIPA_csv : csv
        The combined file with all features. THOIPA prediction is added as a new column
    THOIPA_ROC_pkl : pickle
        Pickled output dictionary with ROC curves
        keys = accessions
        values = dictionary with fpr, tpr etc for each protein
        Could not be saved easily as a dataframe, because the number of residues is different for each protein
    path_uniq_TMD_seqs_for_PSSM_FREECONTACT : str
        Path to text file with list of TMD sequences
    pssm_csv : str
        Path to csv file with the PSSM for the TMD region.

        TMD sequence plus surrounding residues (e.g. 20 each side)
    tm_surr_left : int
        Number of surrounding residues to the N-terminus (left)
    tm_surr_right : int
        Number of surrounding residues to the C-terminus (right)
    motifs_file : str
        Path to csv containing the features related to sequence motifs
    logging : logging.Logger
        Python object with settings for logging to console and file.
    TMD_start : int
        Start of TMD in full sequence
    seqlen : int
        Length of full sequence
		
    entropy_file : str
        Path to csv file with entropy (conservation) data
    pssm_csv : str
        Path to csv file with the PSSM for the TMD region.
    lipo_csv : str
        Path to csv with the lipophilicity features
    freecontact_parsed_csv : str
        Path to csv with coevolution features
    relative_position_file : str
        Path to csv file with the relative position of each residue in the TMD and full protein sequence
    LIPS_parsed_csv : str
        Path to csv with LIPS output organised into features for machine learning.
    motifs_file : str
        Path to csv containing the features related to sequence motifs
    alignment_summary_csv : str
		Path to csv file containing the summary of the alignments and homologues 
		(e.g. how many homologues before and after filtering)
		